List all available agents and workflows in the Azure AI Foundry project.

Returns a formatted list of published agents including their IDs, models, descriptions, and available tools/capabilities.

Example prompts:

• "What agents are available in the project?"

• "List all AI workflows I can invoke"

• "Show me the agents and their capabilities in this Foundry project"

Invoke an agent or workflow with a task and optional context.

Creates a new conversation thread, submits the task, and returns an invocation ID that can be used with agents_get_invocation_status and agents_get_invocation_result.

Check the status of an agent or workflow invocation.

Retrieve the text or file results from a completed agent or workflow invocation.

Create the project log search index in Azure AI Search.

Sets up the index schema including vector search capabilities for semantic similarity search on the context field. The schema supports: title, type, customer_name, short_summary, context (+ embedding vector), project_name, tags, reference_url, architecture, creation_date, modified_date.

Safe to call if the index already exists – it will return a confirmation without modifying the existing index.

Example prompts:

• "Set up the project log search index"

• "Create the Azure AI Search index for storing project summaries"

• "Initialize the vector database schema for project logs"

Ingest a project log entry into the Azure AI Search index with vector embeddings.

Generates a vector embedding for the context field and stores the complete project log entry. Creates the index automatically if it does not exist.

Search the project vector database using semantic similarity.

Generates a vector embedding for the query and returns the most similar documents from the Azure AI Search index.

Add a new document to the project vector database.

Generates a vector embedding for the content and stores the document in the Azure AI Search index for future semantic searches.

List the available sample workflow and agent definitions.

Returns the names, locations, and descriptions of the built-in declarative YAML files in src/foundry_agents/definitions/ and the available CLI deployment/run commands.

Example prompts:

• "What sample workflows are available?"

• "Show me the built-in workflow definitions"

• "List the declarative agent templates I can deploy to Foundry"

Run the full project-log ingestion workflow for a Microsoft customer story.

This workflow sequentially invokes two declarative agents:

1. CaseStudyAgent – fetches the story page and extracts: title, customer name, summary, context, tags, and the source URL.

2. ArchitectureAgent – generates a structured JSON architecture diagram from the case study context and technology tags.

The combined result is stored as a single entry in the Azure AI Search project-log vector index.

If CaseStudyAgent and ArchitectureAgent have been deployed to Azure AI Foundry (via deploy-case-study-agent / deploy-architecture-agent), they are invoked via the Foundry API so that the run is visible in the project telemetry. Otherwise the same logic runs locally against Azure OpenAI directly.